Training dreambooth with multiple classes?

[chetanbanandur]

I'm currently exploring how to expand autotrain-advance to handle multiple classes simultaneously. Specifically, I'm looking to train a model that can recognize two distinct people as well as different environments (like parks or rooms) and then prompt these classes together seamlessly. Are there example implementations or adaptations that could guide me in training across multiple classes? Any insights or suggestions on approaching this would be hugely appreciated. Thanks a lot!